The Box: How the Shipping Container Made the World Smaller and the World Economy Bigger

by Marc Levinson

  An intense national debate ensued. On one side, along with railroads wanting more flexibility to compete with truckers, were shippers and consumer advocates who argued that deregulation would reduce costs. Some trucking companies, especially those that handled smaller shipments, were eager to get rid of regulations. On the other side, many companies that handled full truckloads of freight were bitterly opposed to changes that would encourage partial truckloads, and the unions representing railroad workers and truck drivers fought changes that would weaken union power and eliminate union jobs. The regulators, who were easing regulations slowly and gradually, warned Congress against haste. “Certain shippers command substantial and sometimes overwhelmingly superior bargaining power,” the ICC’s chairman cautioned, asserting the need for the government to keep control in order to protect truckers and railroads from their customers.35

  In the midst of this heated campaign, the container became a poster child for the inefficiency caused by outdated regulation.

  The basic concept of the container was that cargo could move seamlessly among trains, trucks, and ships. Two decades after Malcom McLean’s first containership, though, container shipping was anything but seamless. In principle, a truck line or a railroad could offer an exporter a “through rate” between St. Louis and Spain, but the through rate was simply the published truck or rail rate for that product from St. Louis to a port, plus the published ship rate for that commodity across the Atlantic. Domestically, truckers did not much like carrying containers long distances from the ports because they might well have to haul them back empty; domestic shippers preferred to use conventional trailers, which did not detach from their chassis, rather than detachable containers. Railroads did have a business carrying domestic “piggyback” truck trailers on flatcars, but the service was attractive only for relatively long trips; sending a trailer piggyback the four hundred miles from Minneapolis to Chicago took eighteen to twenty-two hours, against eight hours or so in a truck. Piggyback was often no bargain, either. The railroads set rates high in the forlorn hope that shippers would use boxcars instead, so putting a trailer on a train often cost more than taking it over the road.36

  The railroads were no more aggressive when it came to containers, without the truck chassis and wheels. When Sea-Land and the railroads had discussed a transcontinental container service back in 1967, the railroads asked for three times the price that the ship line was willing to pay, and talks went no further. They tried again in 1972 with a service called “minibridge,” in which a ship line and railroads would join forces to carry a container from, say, Tokyo to New York via the port of Oakland. The carriers would agree on a single rate for the entire trip, file it with both the Interstate Commerce Commission (the rail regulator) and the Federal Maritime Commission (the ship regulator), and decide how to split up the money. Ship lines claimed that minibridge cut costs by eliminating the long, fuel-consuming voyage through the Panama Canal. The real advantage, less publicized, was that loading and unloading ships was much cheaper in Pacific coast ports than on the East Coast: almost no one bothered to export from California to Europe by minibridge through New York. The railroads were so uninterested in the concept that they could not even be bothered to design equipment more efficient than their standard flatcars. Shippers often saw little saving. Sending televisions from Japan to New York by mini-bridge took several days less than by all-water service but was no less expensive. Sending synthetic rubber from Texas to Japan, a U.S. government study found in 1978, cost three times as much by mini-bridge through Los Angeles as when the rubber was trucked to Houston and loaded on a ship.37

  Deregulation changed everything. In two separate laws passed in 1980, Congress freed interstate truckers to carry almost anything almost anywhere at whatever rates they could negotiate. The ICC lost its role approving rail rates, except for a few commodities such as coal and chemicals. Trucks and railcars that had often been forced to return empty were able to get cargo for backhauls. Another definitive break from the past proved critical to driving down the cost of international shipping. For the first time, railroads and their customers could negotiate long-term contracts setting rates and terms of service. The long-standing principle that all customers should pay the same price to transport the same product gave way to a system that yielded huge discounts for the biggest customers. Within five years, 41,021 contracts between railroads and shippers were filed with the ICC. Freight transportation within the United States was reshaped dramatically. Costs fell so steeply that by 1988, U.S. shippers—and, ultimately, U.S. consumers—saved nearly one-sixth of their total land freight bill.38

  Perhaps no part of the freight industry was altered more than container shipping. The ability to sign long-term contracts gave railroads an incentive to develop a business that had languished for two decades, with assurance that their investment would not go to waste. Equipment manufacturers went back to work on low-slung railcars designed for fast loading of containers stacked two-high, the sort of cars Malcom McLean had tried—and failed—to convince railroads to use back in 1967. Deregulation meant that those doublestack cars could be used to haul international containers in one direction and containers filled with domestic freight in the other—impractical before 1980—so the international shipment did not have to bear the cost of returning an empty container to the port.

  In July 1983, American President Lines sponsored the first experimental train composed only of the new double-stack cars. Within months, ship lines and railroads had negotiated ten-year contracts under which dedicated double-stack container trains would speed imports from Seattle, Oakland, and Long Beach directly to specially designed freight yards in the Midwest. Days were shaved off the delivery time. The rates, set by negotiation rather than regulation, were far lower than before and were designed to fall further as volumes rose. On average, it cost four cents to ship one ton of containerized freight one mile by rail in 1982. Adjusted for inflation, that cost dropped 40 percent over the next six years. Rail rates fell so steeply that by 1987, more than one-third of the containers headed from Asia to the U.S. East Coast crossed the United States by rail rather than making the voyage entirely by sea. A major obstacle to international trade had given way.39

  With U.S. trucks and trains deregulated, shipper interests turned their attention to the maritime industry. Once more, they won a sweeping victory. The Shipping Act of 1984 rewrote the rules governing international shipping through U.S. ports. Shippers could now sign long-term contracts with ship lines. In return for guaranteeing a minimum amount of cargo, a shipper could negotiate a low rate and specific terms of service, such as the frequency of ships. These “service contracts” had to be made public, so other shippers with similar freight could demand the same deal. While conferences were still permitted to set rates, individual conference members were free to depart from conference rates whenever they wished, so long as they served public notice.

  Shippers’ newfound power put enormous downward pressure on freight rates. The official rates published by railroads and ship lines did not fall; if the improbable figures in Lloyd’s Shipping Economist are to be believed, the conference tariff for a 20-foot container from Britain to New York doubled between 1980 and 1988. But the official rates meant nothing. A better indication of true market conditions comes from the rates bid for U.S. military freight. The military market was open only to U.S.-flag ship lines, which submitted sealed bids every six months to carry general cargo in containers at least 32 feet long. Ship lines were not obligated to bid, so whatever bids were submitted were presumably above the rates the carriers thought they could earn from commercial cargo. In October 1979, the low bidders offered $40.94 to carry 40 cubic feet of cargo either way across the Pacific. By 1986, the transpacific rates had collapsed to $2.39 westbound, $15.89 from Asia to the U.S. West Coast. Even as U.S. producer prices were rising by nearly one-third between 1979 and 1986, maritime freight rates were plummeting.40

  After the middle of the 1970s, the growth of
nonconference ship lines and the ability of shippers to negotiate rates made official tariff schedules useless as indicators of what exporters and importers were paying to ship their goods. “The rates actually charged vary widely and often deviate substantially from published tariffs,” the World Bank confirmed. The New York Times was less diplomatic, reporting in 1986 that “the shipping world has been turned upside down by five catastrophic years of tumbling freight rates, rising costs, and sinking values of used ships.” The magnitude of the saving to shippers and consumers cannot be calculated, but it was extremely large. When American President Lines studied the matter a few years later, it concluded that freight rates from Asia to North America had fallen 40 to 60 percent because of the container.41

  Chapter 14

  Just in Time

  Barbie was conceived as the all-American girl. In truth, she never was: at her inception, in 1959, Mattel Corp. arranged to make her at a factory in Japan. A few years later it added a plant in Taiwan, along with a large cadre of Taiwanese women who sewed Barbie’s clothes in their homes. By the middle of the 1990s, Barbie’s citizenship had become even less distinct. Workers in China produced her statuesque figure, using molds from the United States and other machines from Japan and Europe. Her nylon hair was Japanese, the plastic in her body from Taiwan, the pigments American, the cotton clothing from China. Barbie, simple girl though she is, had developed her very own global supply chain.1

  Supply chains like Barbie’s are a direct result of the changes wrought by the rise of container shipping. They were unheard-of back in 1956, when Malcom McLean placed his first containers on board the Ideal-X, and in 1976, when high oil prices brought sky-high freight costs that stifled the flow of world trade. Until then, vertical integration was the norm in manufacturing: a company would obtain raw materials, sometimes from its own mines or oil wells; move them to its factories, sometimes with its own trucks or ships or railroad; and put them through a series of processes to turn them into finished products. As freight costs plummeted starting in the late 1970s and as the rapid exchange of cargo from one transportation carrier to another became routine, manufacturers discovered that they no longer needed to do everything themselves. They could contract with other companies for raw materials and components, locking in supplies, and then sign transportation contracts to assure that their inputs would arrive when needed. Integrated production yielded to disintegrated production. Each supplier, specializing in a narrow range of products, could take advantage of the latest technological developments in its industry and gain economies of scale in its particular product lines. Low transport costs helped make it economically sensible for a factory in China to produce Barbie dolls with Japanese hair, Taiwanese plastics, and American colorants, and ship them off to eager girls all over the world.

  These possibilities first drew notice in the early 1980s, when the world discovered just-in-time manufacturing. Just-in-time, a concept originated by Toyota Motor Company in Japan, involves raising quality and efficiency by eliminating large inventories. Rather than making most of its own components, as competitors did, Toyota signed long-term contracts with outside suppliers. The suppliers were intimately involved with Toyota, helping design its products and knowing the details of its production plans. They were required to adopt strict quality standards, with very low rates of error, so that Toyota would not need to test the components before using them. The suppliers agreed to make their goods in small batches, as required for Toyota’s assembly lines, and to deliver them within very narrow time windows for immediate use—hence the name, just-in-time. Keeping inventory to a minimum brought discipline to the entire manufacturing process. With few components in stock, there was little margin for error, forcing every firm in the supply chain to perform as required.2

  The wonders of just-in-time were unmentioned outside Japan before 1981. In 1984, as Toyota agreed to assemble cars at a General Motors plant in California, U.S. business publications ran thirty-four articles on just-in-time. In 1986, there were eighty-one, and companies around the world were seeking to emulate Toyota’s high-profile success. In the United States, two-fifths of the Fortune 500 manufacturers had started just-in-time programs by 1987. Overwhelmingly, these companies found that just-in-time required them to deal with transportation in a very different way. No more would manufacturers offer a load or two to some truck line’s hungry salesman. Now, they wanted large-scale relationships with a much smaller number of carriers able to meet stringent requirements for on-time delivery. Customers demanded written contracts that imposed penalties for delays. Even shipments from another continent were expected to arrive on schedule. Railroads, ship lines, and truck lines with large route networks and sophisticated cargo-tracking systems had the edge.3

  Before the 1980s, logistics was a military term. By 1985, logistics management—the task of scheduling production, storage, transportation, and delivery—had become a routine business function, and not just for manufacturers. Retailers discovered that they could manage their own supply chains, cutting out the wholesalers that had stood between manufacturers and consumers. With modern communications and container shipping, the retailer could design its own shirts and transmit the designs to a factory in Thailand, which used local labor to combine Chinese fabric made from American cotton, Malaysian buttons made from Taiwanese plastics, Japanese zippers, and decorations embroidered in Indonesia. The finished order, loaded into a 40-foot container, would be delivered in less than a month to a distribution center in Tennessee or a hyper marché in France. Global supply chains became so routine that in September 2001, when U.S. customs authorities stepped up border inspections following the terrorist attack that destroyed the World Trade Center in New York, auto plants in Michigan began shutting down within three days for lack of imported parts.

  The improvement in logistics shows up statistically in reduced inventory levels. Inventories are a cost: whoever owns them has had to pay for them but has yet to receive money from selling them. Better, more reliable transport has permitted companies to obtain goods closer to the time they need them, instead of weeks or months in advance, tying up less money in goods sitting uselessly on warehouse shelves. In the United States, inventories began falling in the mid-1980s, as the concepts of just-in-time manufacturing took root. Manufacturers such as Dell and retailers such as Wal-Mart Stores have taken the concept to extremes, designing their entire business strategies around moving goods from factory floor to customer with minimal time in between. In 2004, nonfarm inventories in the United States were about $1 trillion lower than they would have been had they stayed at the level of the 1980s, relative to sales. Assume that the money needed to finance those inventories would have to be borrowed at 8 or 9 percent, and inventory reductions are saving U.S. businesses $80-$90 billion per year.4

  This precision performance would have been unattainable without containerization. So long as cargo was handled one item at a time, with long delays at the docks and complicated interchanges between trucks, trains, planes, and ships, freight transportation was too unpredictable for manufacturers to take the risk that supplies from faraway places would arrive right on time. They needed to hold large stocks of components to ensure that their production lines would keep moving. The container, combined with the computer, sharply reduced that risk, opening the way to globalization. Companies can make each component, and each retail product, at the cheapest location, taking wage rates, taxes, subsidies, energy costs, and import tariffs into account, along with considerations such as transit times and security. The cost of transportation is still a factor in the cost equation, but in many cases it is no longer a large one.

  Globalization, historians and economists have hastened to point out, is not a new phenomenon. The world economy became highly integrated in the nineteenth century. The decline of tariffs and other trade barriers in the years following the Napoleonic Wars led international trade to increase after decades of stagnation, and the introduction of the oceangoing steamship in the 1840s sharply
reduced transport costs. Ocean freight rates fell 70 percent between 1840 and 1910, encouraging increased shipment of commodities and manufactured goods around the world, while the telegraph—the nineteenth-century counterpart of the Internet—gave people in one location current information about prices in another. Prices of grain, meat, textiles, and other commodities converged across borders, as traders found it easy to increase imports whenever domestic prices rose or domestic wages got out of hand.5

  The globalization of the late twentieth century took on quite a different character. International trade is no longer dominated by essential raw materials or finished products. Fewer than one-third of the containers imported through southern California in 1998 contained consumer goods. Most of the rest were links in global supply chains, carrying what economists call “intermediate goods,” factory inputs that have been partially processed in one place and will be processed further someplace else. The majority of the metal boxes moving around the world hold not televisions and dresses, but industrial products such as synthetic resins, engine parts, wastepaper, screws, and, yes, Barbie’s hair.6

  In international production-sharing arrangements of this sort, the manufacturer or retailer at the top of the chain will find the most economical place for each part of the process. This used to be impossible: high transportation costs acted as a trade barrier, very similar in effect to high tariffs on imports, sheltering the jobs of production workers from foreign competition but imposing higher prices on consumers. As the container made international transportation cheaper and more dependable, it lowered that barrier, decimating manufacturing employment in North America, Western Europe, and Japan, by making it much easier for manufacturers to go overseas in search of low-cost inputs. The labor-intensive assembly will be done in a low-wage country—but there are many low-wage countries. The various components and raw materials will come from whichever location can supply them most cheaply—but costs in different locations often are quite similar. Even small changes in transport costs can be decisive in determining where each stage of the process will occur.7